Illustrations by David Foldvari

ONE THING THAT that executives at even the most successful corporations struggle with is how to lure talented candidates into their organizations or professions — how to "incentivize" a job, as an economist might put it — so that an entry-level position has a chance to become a lasting career. The most direct method, of course, is to pay a dazzling salary, a common tactic at many of this country's elite law firms, which now pay first-year associates up to $145,000 a year (not including bonuses). Tech companies like Microsoft and Google, meanwhile, woo engineers with visions of autonomy and stock options. Tighter financial constraints have prompted the social sector, on the other hand, to test different enticements. New York City recently introduced a program of housing subsidies — up to $14,600 annually — available to some teachers who might otherwise look for work elsewhere because of the area's stratospheric housing prices. And a few years ago, the National Institutes of Health began an ambitious experiment when it received Congressional authorization, and many millions of dollars, to create an educational loan-repayment program designed in large part to increase the ranks of this country's "physician-scientists" — that is, M.D.'s who forgo private practice for a career that mixes patient care, typically at a university hospital, with research into diseases, diagnoses and treatments. In the entire United States, there are perhaps only 15,000 physician-scientists, down from more than 23,000 in the mid-1980's. For years, N.I.H. administrators have been watching this decline with concern, as medical-school graduates, carrying debt loads of $100,000 or more, increasingly choose private practices that are far more lucrative than scientific research posts.

This career shift within the medical community might strike those on the patient side as arcane and trifling. But it shouldn't. New cures and treatments don't come from big drug companies alone. They're just as likely to depend on the work of doctors in the field who treat sick patients and discern pathologies that a Ph.D. in a lab cannot see. In the early days of the AIDS epidemic, for example, when patients first presented symptoms of a confounding immunodeficiency disease, it was impossible to start with curative work at the laboratory bench. Doctors working with patients were the ones who tracked the symptoms of the disease and then tested various hypotheses at patients' bedsides. According to Alfred Johnson, a cancer research scientist at N.I.H. who oversees its loan-repayment program, the country's small corps of physician-scientists made one of the biggest contributions to the fight against H.I.V. and AIDS.

Each year since 2002, the N.I.H. has offered to pay off as much as $35,000 of a medical resident's educational debt in exchange for his post-residency commitment to work in a salaried physician-scientist post — whether at the N.I.H., at a university lab or in the field as a researcher. So far, the N.I.H. has accepted nearly 4,000 applicants into its loan-repayment program. The N.I.H. is among the many other organizations, particularly graduate schools and state governments, that have begun using loan forgiveness in recent years to encourage teachers, lawyers, social workers and health professionals to work in underserved geographic areas or among needy populations.

One week in April, I visited some of the nurses and doctors in Minnesota who had agreed to take jobs in the state's far-flung small towns for three or four years in return for a sizable reduction in their student debt — anywhere from $3,000 to $16,000 a year. Last fall, Minnesota extended these loan-forgiveness contracts to rural pharmacists too. Like the physician-researcher, the rural pharmacist is a hidden stitch in the social fabric. "Often these pharmacists will do a kind of triage," Tom Larson, a pharmacy professor at the University of Minnesota, told me. Many work in small towns that lack a doctor, making them the primary health-care expert for miles around. These jobs in the country pay well enough — typically, between $75,000 and $90,000 a year — but they are less lucrative than city jobs, and small-town pharmacists are closer to retirement than their urban counterparts. For Larson and many of his colleagues, something like loan forgiveness is needed to help draw pharmacy grads out into the rural areas, near the farms.

Whether these incentives are financially efficient and socially effective remains an open question among educators and economists. When I asked Alfred Johnson whether the N.I.H. program can replenish the ranks of physician-scientists, he told me that several years of hard data — the N.I.H. is tracking the careers of those who have received loan repayments as well as those who were turned down — are needed before a verdict can be rendered. Even then, it may take a decade or more to evaluate the results. For now, the evidence is anecdotal: e-mail messages from doctors telling Johnson that the loan repayment changed their lives, that it helped them enter a field they never could have afforded otherwise.

A similar indeterminacy colors Minnesota's effort to influence the distribution of pharmacists. Any proof of its effectiveness is years away, but the pharmacists I spoke with equated loan forgiveness with a winning lottery ticket. Their debt was generally about $90,000. Uniformly, they considered it so burdensome as to determine their life choices. "When the lady told me I was accepted to the program," Kurt Schiffler, a young pharmacist in the town of Sauk Centre, told me, "I really thought I was talking with an angel."

O HIGHER TUITION, MORE AND MORE LOANS ne pharmacist's debt is another health administrator's opportunity — in Minnesota, to subtly re-engineer rural society. But whether debt can help an organization recruit candidates, or even, more grandly, whether a wholesale rejiggering of this country's student-loan policies can help ensure that we have enough teachers, pharmacists, nurses, doctors and computer engineers for the 21st century, would seem to depend on how profoundly debt affects a student's (or a graduate's) behavior. In an era when an increasing number of students borrow heavily to finance their educations, are those with less debt, or no debt at all, more likely to stay in school longer? Will they choose a different career path upon graduation? Will they marry earlier, buy a house sooner or save more for retirement? In short, will they live different lives?

A striking aspect of America's educational system is the fact that extensive changes in the financing of higher education over the past three decades have failed to inspire a groundswell of discontent. The transformation has been relentless: as the expense of paying for teachers and research has risen, the cost of educating each student has risen, too. That cost, however, is not the same as what a student pays in tuition. State colleges have long been heavily subsidized by state governments, a factor that has kept tuition manageable and, in many cases, even quite low. Yet these subsidies have diminished over the past couple of decades, forcing students (or parents) to shoulder more of the costs. This evolution can be seen as another form of market privatization. Ed St. John, a professor of education at the University of Michigan, says that the increases in tuition reflect a societal choice to make the "consumers" of education carry more of its costs. Or as he puts it: "We've shifted taxation for education from taxpayers to individuals in the form of loans." In St. John's view, the free-market transition has produced benefits: hundreds of small colleges that otherwise might have been expected to close have flourished instead, and the percentage of high-school students who go on to college has risen steadily. At the same time, however, complex financing schemes have become a fact of life for students and their families. Those from middle- and lower-income backgrounds for whom financial aid falls short — federal grants have not kept up with tuition costs — have taken on loans that are more likely to be onerous and, in many instances, life-altering.

This is an unintended consequence. Since the early 1970's, the college-aid system — now comprising federal programs like Pell Grants, Perkins Loans, Stafford Loans, PLUS loans for parents and myriad state and institutional programs — has been outfitted with countless gears and pulleys. "It's like the tax system," Sandy Baum, an economist for the College Board who teaches at Skidmore College, says. "Each piece gets piled on another piece. And the way they fit together is generally not something people would design by purpose." Even simple questions like what's the best way to save for college — a 529 college savings plan? a Coverdell Education Savings account? — have become incredibly confusing.

One thing that has become easier, though, is borrowing. And not just small sums. Two-thirds of all graduates of four-year colleges now finish with sizable loans; their average burden, according to the Project on Student Debt, a task force financed by the Pew Charitable Trusts, has risen by more than 50 percent over the past decade, even after accounting for inflation. With respect to public colleges, which turn out a majority of United States graduates annually, those who borrow now finish with a median of about $16,000 of debt. The figure for a private school graduate who borrows is greater — about $20,000.

Many educators believe that these numbers portend an emergency. They also view undergraduate debt as materially different from — and more threatening than — graduate- or professional-school debt, which ranges between $50,000 and $150,000 a student. That's because the earning power of an M.B.A. or a medical or law degree is far greater. "When young people are training to be doctors or lawyers they should pay the freight because they can earn so much more," says Elizabeth Warren, a professor at Harvard Law School and a leading expert on personal bankruptcy. On the other hand, Warren is convinced that rising levels of undergraduate debt have diminished the potential — long taken for granted — of a four-year degree to deliver someone into the middle class. "We tend to talk about student loans in the abstract, 'Ten or twenty thousand dollars — it's not that much,"' she explains. "But I think it's really about what it means to be 28 and try to make loan payments and health insurance premiums and still put something aside for a down payment for a house. Think about how much extra room you have to have in your budget to cover those three things. Most can't do it." When Warren was at the University of Houston in 1970, she notes, her tuition was $52 a semester. Skip to next paragraph Related Is there a student-loan crisis in America, and if so, what should be done about it? Readers’ Opinions Forum: Forgive Us Our Student Debts

The view that today's student debt may have explosive consequences for society seems to be gaining support in some quarters. Recently, two authors, Tamara Draut and Anya Kamenetz, have made forceful arguments in their books ("Strapped" and "Generation Debt," respectively) that college debt — and particularly the repayment arrangements between the federal government, private banks and students — interfere with the dreams and potential of an entire 20-something generation. "It's not about whether these policies are stopping these people from applying to college, though there is some evidence that they are," Kamenetz, a 2002 Yale graduate, says of the rising borrowing rates. "It's whether they're fair, whether they represent our ideals as a society and whether in the long term they're effective as compared to the policies of other countries that are pulling ahead of us."

One day this spring, Susan Dynarski, an associate professor of public policy and an economist at Harvard, joined Kamenetz in a colloquy at the American Enterprise Institute to address the question of whether college is still worth the cost. In conversation, Dynarski often returns to an incontrovertible point: a bachelor's degree remains a tremendously good investment, and virtually all estimates put its value at a million dollars in extra earnings over the course of a lifetime. "It's never been a better time to be a college graduate and a worse time to be a high-school dropout," Dynarski says. While she agrees that student borrowing has risen steadily, she counters that critics like Kamenetz need to put that debt into perspective: "Credit card debt is up, car debt is up, people are spending more," she says. "Among all of those sources of debt, a college education is one of the best reasons to go into debt." For those who get a bachelor's degree at a public university and end up owing the median of $16,000, Dynarski puts their monthly payment at about $210 a month. Meanwhile, she points out, average car debt for people who borrow money to buy a new car is $24,000, with a monthly payment of about $450.

The issue she raises seems clear enough: Is the problem really student debt? Or is it that student debt is now colliding with a larger culture of consumer debt? And if that's the case, which takes precedence?

M THE BORROWERS ost of the questions researchers ask about the links between college borrowing and behavior focus on what happens when students from low-income backgrounds are faced with high tuition and inadequate financial aid. In academic circles, this is called the "work-loan burden": the amount of money students have to come up with out of their own pockets to be able to attend school. Faced with a high work-loan burden, do they work for another year before attending college? Or do they go to a cheaper, two-year school instead of a more expensive four-year one? Should they study part time and work part time? For low-income high-school seniors, it's conceivable that virtually every decision made about college is influenced by their willingness to borrow. According to one recent estimate, 200,000 qualified students every year put off college for financial reasons.

"I think the key issue is that these students' aversion to debt is a rational choice," Nicole Barry, deputy director of the federal Advisory Committee on Student Financial Assistance, says. "If work-loan is a high percentage of their family income, then it makes sense that they would be averse to taking on such high levels of debt." As Barry also points out, one of the most punishing outcomes occurs when those with a high work-loan burden take out loans to pay for their tuition for a year or two but never complete their degree, saddling them with a future of monthly payments but little chance of reaping any benefit from their investment. A recent report for the National Center for Public Policy and Higher Education, a nonpartisan research organization, estimated that about one in five students who take on loans fall into this category.

To Dynarski of Harvard, low-income students should be our real worry, rather than middle-class students with debts. A few years ago, Dynarski calculated that $1,000 in grant aid increases the likelihood of a student attending college by about 3.6 percentage points. Since then, she has written several papers that consider the effects of various state merit scholarships that reward high school students who maintain, say, a B average or better with partial or full college tuition. Dynarski discovered that these programs typically increase the probability that students will attend college by 5 to 7 percentage points. What's more, they shift students toward four-year schools and away from two-year schools. Merit-aid recipients are slightly more likely to finish their degrees, Dynarski told me. While Georgia's HOPE scholarship has stirred controversy for having seemingly been awarded disproportionately to those from affluent backgrounds, Dynarski defends merit scholarships. The bottom line, she says, is that "these programs increased college enrollment quite a bit."

Other experiences have yielded similar conclusions about costs and access. In 1999, the Bill & Melinda Gates Foundation established the Gates Millennium Scholarship Program, a yearly award given to 1,000 high-achieving students of color from low-income backgrounds. The scholarship is what's known as a "last dollar" scholarship — meaning that once a student has a final offer of financial aid from a college, the Gates Foundation will cover all remaining costs. In effect, it reduces a student's work-loan burden to zero. The scholarship covers up to five years of school and may extend through some graduate schooling as well. When I spoke with Jim Shelton, who is the foundation's program director for education, he told me that the scholarship has two goals: to make a difference in recipients' lives and to help them become leaders in their communities. "By reducing their financial barriers," Shelton says, "you're allowing students to take on leadership roles on college campuses and spend more time focused on academics. Otherwise they would have been burdened with work or making lesser college choices."

The foundation intends to track some of the scholarship winners though age 36, but last year, researchers sifted through the program's data for the first graduating class. What they found, Shelton explains, is that when the financial barriers were removed, students became far higher achievers. A student in the program who would have tended to go to a two-year school went to a four-year school instead; a student who would have chosen a fairly nonselective school chose a very selective school. And students who previously might have gone to public schools opted for more elite private schools. Another curious effect emerged, too. Ed St. John, the Michigan professor and a researcher working with the Gates Foundation, noticed that the Millennium scholars, compared with a control group, are more likely to take a job related to their college major and that more frequently they are choosing education. "The people who don't have to worry about debt, like the Gates cohort," he says, "are making life choices that are more contingent with their interest rather than to the market." Put simply, among low-income high-achievers, those who don't have to take out loans often make different life choices from those of their counterparts who borrow.

B THE PSYCHOLOGY OF DEBT etween 1987 and 2002, the student lender Nellie Mae surveyed American students four times to get a sense of how much debt they carry and how it affects their professional outlooks. Sandy Baum, an economist in charge of the surveys, told me that over the years there has been a slight "creeping up" of the correlation between debt and the pressure felt by undergraduates to go into certain careers. Still, as recently as four years ago, less than 20 percent of those surveyed said their loans influenced their postgraduate career plans. Baum agrees that the percentages may be higher today, in light of tuition increases; she also says that an increase in college-loan interest rates this summer, which may push up monthly payments for millions of students, makes her think that more families will be struggling. And yet Baum is very much in the rationalists' camp, noting that a degree remains an extremely good investment and that student debt is not yet a full-blown crisis. "If you graduated from college, and you could be a social worker or an investment banker, the fact that you have $20,000 in student debt is hardly a deciding factor," she says. "It would be irrational to make a lifetime decision based on that."

Maybe. But students don't necessarily behave rationally, especially when it comes to money. (Mounting evidence in behavioral economics shows that the rest of us don't, either.) Moreover, what's sensible for an upper-income family may seem unconscionably risky to the less affluent. Consider Princeton's experience. When the school replaced all student loans with grants five years ago, it apparently created the first modern university class in the United States — the class of '05 — without any significant debts. "What we were after in eliminating loans was to make it simpler for students of modest means to think about private higher education," Nancy Malkiel, dean of Princeton's college, says. "Those families, we thought, based on what we were hearing, couldn't quite imagine taking out loans. It wasn't part of their experience, it wasn't part of their idea of what was sensible and feasible." The administrators, Malkiel adds, did not intend to influence students' post-college careers; mainly they wanted to ensure that a high work-loan burden would not drive lower- or middle-income students away. Still, when the financial aid office surveyed the class of '05 just before graduation last spring, it learned that two-thirds of the students sampled said that graduating with little or no debt had made a significant impact on their postgraduate plans. Some said that becoming a teacher, or working for a nonprofit, was now a viable option.

This is particularly interesting in light of a White House report that predicted that over the next few years the United States will need to recruit two million new teachers for the nation's classrooms. The mathematics of student debt makes such a goal seem utopian. Two months ago, the State Public Interest Research Groups' Higher Education Project calculated that student-debt loads will be unmanageable for 38 percent of private-school graduates who enter the teaching profession and for nearly 55 percent of private-school graduates who become social workers. (At public four-year colleges, the numbers were 23 percent for teachers and 37 percent for social workers.) While Congress added provisions in 2004 for debt forgiveness on new student loans to help recruit more math, science and special ed teachers to underserved areas, and while a number of states (and cities like New York) have added various incentives to do the same, these steps may not address the crux of the teacher-shortage problem. Nor will easier licensing standards. If we want more teachers, we'll probably need to try something financially drastic.

That's because loan forgiveness can't necessarily change how students think about debt and careers right after graduation. Perhaps the best research on this subject comes from the experience of several hundred recent New York University School of Law graduates. The pay disparities between public- and private-sector law jobs have widened over the past two decades to the point that law-firm jobs can pay three to five times what graduates earn in the public realm. Many elite law schools, and many bar associations and public interest law groups, have responded to the disparity with various pay incentives — mainly, debt-forgiveness programs that try to ensure that highly qualified lawyers will continue to take jobs as public defenders and in law enforcement. But from 1998 to 2001, N.Y.U. Law tried something called the Innovative Financial Aid Study. Of the participants who received aid, some students received loans (meaning they would graduate with large debts) and some received grants of two-thirds tuition (meaning they would graduate with smaller debts). This randomized arrangement came with conditions, however: students who received loans could have them forgiven if they took jobs in the public sector; students who received grants would see them converted into loans (and large debts) if they did not enter the public sector. N.Y.U. is tracking these students' job choices for 10 years.

For all of them, of course, the choice was the same: low debt if they took jobs in the public sector versus high debt if they went into the private sector. But the psychological difference between completing school with low debt (thanks to tuition grants) or high debts (albeit with a promise of loan forgiveness) proved significant. When Erica Field, an assistant professor of economics at Harvard, analyzed the early survey data, she concluded that grant recipients had a strikingly higher rate (between 36 and 45 percent) of first-job placement in public sector jobs than those who had received loans instead. "I think what my results suggest is that people are debt averse and there's a psychic cost to debt," Field says. "If you can reduce the amount of time that they're in debt, people will value this. You can influence their behavior."

The problem, and it's a big one, is not only that college can't be made free for everyone. It is also getting more expensive every year. N.Y.U.'s immense resources enabled it to explore the effects of big scholarships on schooling; only a handful of other institutions in the world can do that. To Jim Shelton at the Gates Foundation, the expense of the Millennium Scholarship Program (even with its $1 billion endowment) is such that he's certain the college-financing problem will never be resolved through private or philanthropic funds. Government must lend a hand. "The reality is that if we really want to address this issue, it will have to be at a public policy level," he says. Skip to next paragraph Related Is there a student-loan crisis in America, and if so, what should be done about it? Readers’ Opinions Forum: Forgive Us Our Student Debts Elizabeth Warren, the Harvard law professor and bankruptcy expert, has spent the spring working out a policy paper that essentially calls for using loan forgiveness to make college free to anyone. "Let's lend any student the money for college," she says, "and then provide four years of loan forgiveness for public-service work. That means that people could be 26 years old, have a college education, four years of work experience and be debt free." In Warren's opinion, the country's main hope to remain competitive is to get more people into college and more students out of debt. "We need a big program, otherwise it's not transformative," she says. "We don't need changes at the margin."

W THE INCOME-CONTINGENT REPAYMENT PLAN hat makes student loans more problematic than other kinds is that the lender is helping someone invest in his own intellectual development, not in a house or small business. But investing in human capital is risky. A lender can't be sure that you'll land a good job after graduation; even worse, it can't secure any collateral. Nor can a bank repossess your skills. More than 50 years ago, the economist Milton Friedman called these aspects of student debt "an imperfection in the capital market"; Friedman suggested that it would be more efficient to tax students a percentage of their future earnings rather than to have them take out fixed-rate loans for their education. His idea has taken many different iterations since then, but income-contingent repayment, as it's now called, has never really gained a foothold in the United States beyond a limited option available in the federal direct-lending program. It is, however, being put fully into practice in several countries, most notably Australia and New Zealand. While the specifics vary, the loans generally allow students to pay nothing while they are in school and to pay a percentage of their income — usually about 4 to 8 percent — after they graduate. Thus a low earner (a new teacher or social worker) pays the same proportion of his income as a high earner.

One main ambassador for an income-contingent repayment system in this country is Thomas J. Kane, who recently moved from U.C.L.A. to Harvard's Graduate School of Education. I met with Kane on a recent afternoon in downtown Manhattan. Kane's message is that this country needs to figure out a better way to pay for college and that income-contingent repayment is simply the most practical solution available. As government subsidies for higher education continue to shrink relative to the costs of running institutions of higher learning, we can either squeeze the quality of our colleges or raise tuitions. "But we can only do that so much," Kane told me. "It just seems like the current system is untenable." He also notes that it makes little sense to base all federal aid on a student's present financial circumstances. Rather than relying only on "backward-looking means testing," as he calls the current criteria for student financial aid, "why not provide some of it based on forward-looking means testing, based on the way your career turns out?" he asks. Income-contingent repayment promises to do that. Who you are — rich, poor, middle class — when you apply for aid before college would not be the main determining factor. And what you become after college wouldn't matter greatly, either. You would still just pay a percentage of your income.

Lately, the word "Australia" has almost become a tic among economists who study educational debt issues, an example of both a big idea and a reasonably good idea. Recently I spoke to one of that system's main architects, Bruce Chapman, an economist at the Australian National University. At the moment, the base line for paying back college debt there is $36,200 (Australian), Chapman told me, which translates into about $27,000 (United States). Any student who makes above that threshold must pay back a portion of his income to the government. There is no monthly check to write; the money is deducted from your pay — much like the Social Security tax here — and is funneled through the tax system. Chapman sees two big advantages to the process. First, a graduate cannot default on an income-contingent loan because of inability to pay. And second, the system offers insurance against hardship in repayment. "Imagine you have a U.S. loan and I have an Australia loan," he says, "and then we both have an accident, or we lose our job, or we're just unlucky for a year. Something that decreases our income substantially. You still have to pay the bank loan. Your income may fall by 70 percent, but the bank doesn't care about that, nor should it, since it's profit-maximizing. In the Australian system, if something unfortunate happens, then the payment is zero." Skip to next paragraph Related Is there a student-loan crisis in America, and if so, what should be done about it? Readers’ Opinions Forum: Forgive Us Our Student Debts

There's no guarantee that a United States system would work like Australia's, of course — there are great differences in the ways the two countries pay for their higher education — but that country's example presents some intriguing possibilities. Kane has worked out a proposal whereby income-contingent loan repayment could be implemented through our tax system without costing more than some current federally financed college-aid programs. Kane says he believes that it would make the gamble of going to college far less risky, attracting new students who might otherwise be debt averse, while removing the inhibitions of recent graduates who want to take lower-wage jobs that they find more interesting than higher-wage ones. I asked Ed St. John at Michigan about this. With a massive income-contingent plan, would we have more teachers and nurses? "I think we would, sure," he said.

S THE SOCIETAL BENEFITS OF A COLLEGE EDUCATION till, to convert the student-aid system into a vast loan-repayment organization as Elizabeth Warren proposes, or to take an arguably more pragmatic approach like Thomas J. Kane's income-contingent scheme, requires more than a group of scholars agreeing that students and families face a debt crisis. Our conception of what college is supposed to do for individuals and society has to be transformed. One remarkable thing about the rise in college tuition and student debt is that it has coincided almost precisely with a perceptual (and rhetorical) shift in the value of a college degree. When young writers like Anya Kamenetz point to the effects that college debt may have on reducing this country's economic competitiveness, they are largely referring to what economists call the "externality" effects of a college-educated public. Jargon aside, this expresses the notion that a college degree doesn't merely yield a private economic benefit (a million dollars over a lifetime) but that an educated public can keep health care costs down (college grads take better care of themselves), raise levels of economic development (they create more jobs and companies) and increase tax receipts (they make more money and pay more taxes).

Efforts to call attention to these external, public benefits of higher education have not entirely withered. Though it often goes unremarked, many of the new merit programs started by various states, like Georgia's HOPE scholarship, arise from a conviction that it is in a state's best economic interests to expand the benefits that accrue to a more highly educated population. In 2004, the College Board issued a lengthy report, written by Sandy Baum and Kathleen Payea, that tries to measure some of those benefits; the report details the civic windfalls of education, and it notes that college graduates smoke less, give blood more and volunteer far more frequently than high-school graduates. (They also have far lower rates of incarceration and rely less on government entitlement programs.) If we want a better, richer, healthier society than we have today, the evidence suggests that the best way to get there is to have more college graduates.

Yet almost certainly the argument about the personal financial payoff of college has become the more compelling narrative — and when the benefits of college are increasingly considered personal, large personal debts sound like a fair trade-off. Susan Dynarski at Harvard suggests that the shift from "public good" to "personal gain" reflects how the rewards of a degree have changed since the 1970's, when the gulf between the earnings of high-school graduates and college graduates was not so substantial. "And then it started going up, up, up," she says of the payoff. In the meantime, the public benefits of higher education — like more competitiveness in the global economy, lower health care costs — have proved extremely difficult to measure. Economists are often left with correlations (more college graduates volunteer, for instance) rather than causations (graduates were inspired in college to volunteer). It can be vexing to calculate the externality effects of college graduates on a small town, city or state, never mind trying to figure out the imprecise impact on our huge national economy.

Loan-repayment programs like those at the National Institutes of Health or in Minnesota are beset with similar limitations. They can't necessarily demonstrate, in numbers, what they do for residents who suddenly get a new local doctor or pharmacist. How do we value the work of physician-scientists dealing with a viral outbreak? Or the casual exchange between a pharmacist and a rural neighbor who describes a pain running down his left arm?

Still, the programs do present a chance to see, in a modest but visible way, that a degree yields a mix of personal and public payoffs. One crisp April morning in Minnesota, I drove north several hours from Minneapolis, past farm fields and frozen lakes just beginning to thaw along their embankments, to a small town of 3,040 called Long Prairie. Kathy Smith, a 41-year-old registered nurse, met me at the Long Prairie Nursing Home, a clean, bright, low-slung complex on a hill at the edge of town. Anyone working in or intending to work in a rural nursing home can apply to Minnesota's loan-forgiveness program. Smith, who grew up in the neighboring town of Browerville, had $22,000 in debt and a commitment to staying in a rural area. She applied and won the forgiveness grant, which works out to about $3,200 a year. "It's the small-town life, that's why I stay here," she told me as we toured one of the home's residential wings. "When you leave work, if you're downtown, if you're at the church, wherever you are, you're still around the people that you've helped and taken care of all day."

Smith did well in high school, she says, but neither of her parents went to college, and no one in her immediate family (there were 11 children) was interested in a degree. "The money was tight, and my parents didn't really have anything when we were in high school, so they couldn't help at all," she said. When she couldn't get a grant after high school, she worked for five years in a factory that makes eyeglasses. Eventually, she earned a license to work as a practical nurse and, at the urging of her bosses at the nursing home, decided to go back to school for two years to become a registered nurse. She drove to classes an hour away in Detroit Lakes — it took much longer when she had to wait at a crossing for a freight train — with a friend in the nursing program. They studied on the drive to and from school.

What seems surprising is that she made it through at all. She'd been baffled by the loan forms and nearly paralyzed by the idea of taking on large debts. "A lot of people, they ask, 'What did it cost you to go to college?"' Smith said. "I say: 'O.K., was that just for the costs of the tuition and the books, or do you want to factor in that we've had to lose hours of work because my husband had a part-time job and I had a part-time job, so both of us had to give those up when school started? And do you factor in the fact that you're not there to take care of your son anymore so now you have day-care expenses?"'

Smith makes $22 an hour today. "I wish I had gone on right away instead of waiting," she told me. "Even if I couldn't afford it back then, I should have. It was a lot of years wasted." For Long Prairie, too, you might argue. If things had turned out different, the town would have had another registered nurse 20 years earlier. And what small differences could that have made?

Jon Gertner, a contributing writer for the magazine, often writes about business.

Copyright 2006 The New York Times Company. Reprinted from The New York Times Magazine of Sunday, June 11, 2006.